The Role of the MiR-181 Family in Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is the fourth-leading cause of cancer-related death worldwide. Due to the high mortality rate in HCC patients, discovering and developing novel systemic treatment options for HCC is a vital unmet medical need. Among the numerous molecular alterations in HCCs, microRNAs (miRNAs) have been increasingly recognised to play critical roles in hepatocarcinogenesis. We and others have recently revealed that members of the microRNA-181 (miR-181) family were up-regulated in some, though not all, human cirrhotic and HCC tissues—this up-regulation induced epithelial–mesenchymal transition (EMT) in hepatocytes and tumour cells, promoting HCC progression. MiR-181s play crucial roles in governing the fate and function of various cells, such as endothelial cells, immune cells, and tumour cells. Previous reviews have extensively covered these aspects in detail. This review aims to give some insights into miR-181s, their targets and roles in modulating signal transduction pathways, factors regulating miR-181 expression and function, and their roles in HCC.


Introduction
Hepatocellular carcinoma (HCC) is the fourth-leading cause of cancer-related death worldwide [1][2][3].More than 50% of patients are diagnosed with advanced HCC.Unfortunately, only a few first-line drugs are available for patients with advanced HCC, such as sorafenib, lenvatinib, and the combination of atezolizumab and bevacizumab.Due to the high mortality rate and unmet treatment options in HCC patients, discovering and developing novel systemic treatment options for HCC is a vital unmet medical need [4][5][6][7][8][9].
Among the numerous molecular alterations in HCCs, microRNAs (miRNAs) have been increasingly recognised for their crucial roles in hepatocarcinogenesis.After being extensively studied over the past two decades, the precise significance and function of the miRNAs in HCC formation and progression remain elusive [10][11][12][13].We and others have recently revealed that members of the microRNA-181 (miR-181) family were up-regulated in some, if not all, human cirrhotic and HCC tissues.Up-regulated miR-181 induced hepatocyte epithelial-mesenchymal transition (EMT), promoting HCC progression [10,[14][15][16][17][18][19].MiR-181 has been widely reported to play vital roles in governing the fate and function of endothelial cells, fibroblasts, and immune cells [10,12,[20][21][22][23][24], which are all involved in cancer initiation and progression to a certain extent.Previous reviews have extensively covered these aspects in detail.Herein is a concise review of miR-181, its targets and regulations, and especially its overlooked significance in HCC.

Database of miRNA Targets
MiRNAs primarily regulate gene expression by targeting gene transcripts.Understanding the potential function of a specific miRNA often begins with exploring its targets.Several miRNA target prediction tools have been developed, but due to the complex mechanism underlying miRNA action, the predicted targets often vary [46,47].Some examples of widely used miRNA target prediction tools are (i) TargetScan (v8.0):TargetScan is a popular miRNA target prediction tool that incorporates information about miRNA seed regions and evolutionary conservation to predict potential target sites [40,48]; (ii) miRanda (v3.3a): miRanda utilizes a scoring algorithm that considers sequence complementarity and conservation to predict miRNA targets [49]; (iii) PicTar: PicTar predicts miRNA targets by integrating sequence complementarity, conservation, and the presence of miRNA target site clusters [50]; (iv) RNAhybrid (v2.2.1): RNAhybrid employs a thermodynamic model to predict the energetically most favourable interaction between an miRNA and its target site [51]; (v) PITA: PITA (miRNA target prediction by base pairing probability) predicts miRNA targets by calculating the base pairing probability between miRNAs and potential target sites [52]; (vi) DIANA-microT (v2023): DIANA-microT employs a machine learning approach to predict miRNA targets by integrating multiple features, such as site accessibility and seed region conservation [53]; and (vii) miRDB (v6.0): miRDB is an online database that provides miRNA target predictions based on a machine learning algorithm trained on experimental data [54].While TargetScan, miRanda, and DIANA-microT are popular and perform well, their false positive and false negative rates are still very high.
Thus, experimental evidence is required to validate miRNA-target interactions [55].MiRNA research has yielded a growing number of experimentally validated miRNA-target interactions since 1993.These valuable findings have been compiled and summarized in several databases, including miRWalk (v3) and miRTarBase (v9.0 beta) [56,57].The miRTarBase is a fully manually curated online database that stores validated miRNA targets.The more validated experimental data that accumulate, the more miRNA targets will be verified.
It should also be noted that phenotypic changes of disrupting single miRNA-target interaction are often subtle for the following reasons: (i) most (>90%) of gene transcripts are targeted by more than one miRNA, so most miRNA targets would be down-regulated by less than 50%; (ii) the protein expression of most miRNA targeted genes can vary by two-fold without causing detectable consequences; and (iii) miRNA targets are regulated by many regulatory elements, i.e., buffering effects of the regulatory network [32].

MiR-181 Targets and Their Roles in Modulating Signal Transduction Pathways
According to TargetScan analysis, it has been predicted that 1371 human transcripts harbour conserved miR-181-5p binding sites.These transcripts collectively contain 1694 conserved binding sites and 960 poorly conserved binding sites.On average, each transcript possesses approximately two binding sites.In the miRTarBase database, there are 2556 documented instances of miR-181, including 1350 validated miR-181 (both miR-181-5p and miR-181-3p) targets.After comparing the predicted miR-181 targets with the validated miR-181 targets, only 435 of the targets overlapped.This discrepancy suggests the limitations of miRNA target predictors due to the intricate mechanisms governing the actions of miRNAs.Interestingly, a new study found an alternative seed match and identified a distinct set of miR-181 targets within the coding sequence of genes, revealing that miR-181 acts primarily through RNA destabilization and translational inhibition [58], but more studies are needed to determine whether the results of this study can eliminate the discrepancy mentioned above.
It has been reported that the expression level of miR-181 is up-regulated by Wnt/βcatenin signalling, which plays a vital role in the pathogenesis of multiple cancers, including HCC [13,16,118].It has been reported that the promoter of miR-181a/b2 contains several TCF/LEF binding sites [118].Interestingly, miR-181s target NLK, TIMP3, and GSK3β, all of which are negative regulators of Wnt/β-catenin signalling [16,74,122], indicating miR-181s and the Wnt/β-catenin signalling may form a positive feedback loop.
In 2009, it was reported for the first time that the expression of the miR-181 family was notably elevated in HCC, typically in liver cancer stem cells [16].Its elevation played a pivotal role in sustaining the population of liver cancer stem cells by inhibiting CDX2, GATA6, and NLK [16].Interestingly, TGF-β was shown to induce the upregulation of miR-181b, which in turn promoted HCC formation by down-regulating TIMP3 [19].We found that miR-181a is one of the most amplified miRNAs during TGF-β-induced hepatocyte epithelial-mesenchymal transition (EMT) and is upregulated in fibrotic liver tissues and HCC samples from humans and mice [14].
The importance of miR-181a in human HCC was further enhanced by a seminal study using TCGA-LIHC human samples, where miR-181a was expressed at much higher levels than any other member of the miR-181 family.This paper grouped HCC cases based on multiple molecular parameters, including genomic mutations, DNA methylation, DNA copy number, and mRNA and miR expression.HCC cases were divided into three clusters, i.e., iClusters 1, 2, and 3. Patients in iCluster 1 had the worst prognosis.Strikingly, miR-181a was highly expressed in iCluster 1 compared with iClusters 2 and 3 [60,171].
Global or conditional miR-181 knockout (KO) mice have been generated and made available to the public.They have been widely used to study the roles of miR-181 in immune cells [20,39,41,44,60,62,93,100,101,108,121,146,[172][173][174][175].These valuable genetically engineered mouse models (GEMMs) are just beginning to be used to study miR-181s in cancer [60,121].Our subsequent experiments using these GEMMS firmly corroborated the vital role of miR-181a/b1 in promoting chemically induced hepatocarcinogenesis [60].In our experiments, liver tumour size was significantly reduced by 90% in global KO (GKO) or liver-specific KO (LKO) mice of miR-181a/b1 but not in hematopoietic and endothelial lineage-specific KO mice.However, we also showed that tumour induction requires both hepatic and non-hepatic miR-181 overexpression [60].We showed that EMT was partially reversed in GKO tumours, linking to earlier studies that showed that miR-181a induced hepatocyte EMT [14].The results obtained from these GEMMs indicate that miR-181a/b1 in tumour cells, but not in tumour-associated cells, plays the dominant role in the formation of DEN-induced liver cancer [60].
According to the results of our animal experiments and analysis of human HCC datasets, CBX7 was clearly shown to inhibit (rather than enhance) liver cancer progression via targeting multiple genes, such as WNT10a, DUSP4, FGFR2, and CCNE1 [60].
In the analysis of these data, we also showed that low miR-181 combined with high CBX7 expression in HCC was associated with the best prognosis and vice versa, with high miR-181 and low CBX7 HCC having the worst prognosis [60].Furthermore, gene expression profiles of miR-181a/b1 KO HCCs in mice overlapped with low-proliferative periportal-type human HCCs with a good prognosis [60,192].Thus, miRT-181a/b1-deficient liver tumours may be an ideal model for low-proliferative periportal-type human HCC.

The miR-181 Family in HCC: Summary and Future Perspectives
In summary, the biological functions of miRNAs and their diagnostic, therapeutic, and prognostic value in liver cancer have been extensively studied to a certain extent [16,18,59,60,135,136,147,149,171].MiR-181a/b1, especially miR-181a, plays a crucial role in the formation of liver cancer by directly controlling the fate and function of tumour cells [12,20,21,60,[71][72][73]. MiR-181s play a crucial pathogenic role in the formation and maintenance of stem cells, including liver cancer stem cells, deserving more studies to explore this aspect [16,17,45,72,99,144,157].It will be very helpful to further thoroughly examine the expression levels of miR-181 members in different subtypes of HCC and different components of TME in humans.To do that, novel technologies are desired to effectively dissect the heterogeneity of HCC, the complexity of the miR-181 family, and their interactions.The emerging technologies, such as single-cell and spatial multi-omics technologies, including miRNA or miRNA-mRNA co-sequencing, may bring more clarity to miR-181s in human HCC [193][194][195].
Among the signalling pathways modified by the miR-181 family [10,11,13,20], TGFβ-induced EMT signalling has been extensively explored and confirmed to be regulated by the miR-181 family.However, the detailed molecular mechanisms responsible for miR-181-induced EMT are largely unknown.CBX7 also plays a vital role in the regulation of EMT [60,76,185,186,[196][197][198].It is unclear how EMT is regulated by the TGF-β-miR-181-CBX7 axis if validated, and the interaction and regulation between TGF-β, miR-181, CBX7, and EMT is likewise unclear (Figure 2).Answering these questions will deepen our understanding of this axis and may provide new targets for intervention.
Once a significant miRNA and its critical targets are identified to play critical roles in the progression of HCC, several approaches can block miRNA activity, such as anti-miRs and blockmiRs [59,135,[199][200][201]. Anti-miRs are antisense oligonucleotides with complementary bases that pair with the miRNA target, leading to miRNA silencing, mainly by sterically blocking the target miRNA [201].One miRNA can target many, even hundreds, of mRNAs.Thus, it is very challenging for the anti-miR to prevent one specific mRNA only from miR-induced inhibition, as it will target many other mRNAs.BlockmiRs use antisense technology to reduce miRNA activity by specifically binding to the target mRNA rather than miRNA [59,201].Whether it can be used for miR-181 and its specific targets is yet to be determined.
181-induced EMT are largely unknown.CBX7 also plays a vital role in the regulation of EMT [60,76,185,186,[196][197][198].It is unclear how EMT is regulated by the TGF-β-miR-181-CBX7 axis if validated, and the interaction and regulation between TGF-β, miR-181, CBX7, and EMT is likewise unclear (Figure 2).Answering these questions will deepen our understanding of this axis and may provide new targets for intervention.Once a significant miRNA and its critical targets are identified to play critical roles in the progression of HCC, several approaches can block miRNA activity, such as anti-miRs and blockmiRs [59,135,[199][200][201]. Anti-miRs are antisense oligonucleotides with complementary bases that pair with the miRNA target, leading to miRNA silencing, mainly by sterically blocking the target miRNA [201].One miRNA can target many, even hundreds, of mRNAs.Thus, it is very challenging for the anti-miR to prevent one specific mRNA only from miR-induced inhibition, as it will target many other mRNAs.BlockmiRs use antisense technology to reduce miRNA activity by specifically binding to the target mRNA rather than miRNA [59,201].Whether it can be used for miR-181 and its specific targets is yet to be determined.

Figure 1 .
Figure 1.Chromosomal position of the three miR-181 clusters.(I) Mir181A/B1 and (II) mir181A/B2 transcription start sites (TSS) have been mapped to 78.3 kb and 34.0 kb upstream of the mature miRNAs, respectively, consistent with the position of H3K27Ac (in blue/pink), which is associated with the higher activation of transcription [38].(III) The putative TSS of miR181C/D might be 9 kb upstream of the miR-181C/D precursor.Please note that the clusters are not drawn to scale.

Figure 1 .
Figure 1.Chromosomal position of the three miR-181 clusters.(I) Mir181A/B1 and (II) mir181A/B2 transcription start sites (TSS) have been mapped to 78.3 kb and 34.0 kb upstream of the mature miRNAs, respectively, consistent with the position of H3K27Ac (in blue/pink), which is associated with the higher activation of transcription [38].(III) The putative TSS of miR181C/D might be 9 kb upstream of the miR-181C/D precursor.Please note that the clusters are not drawn to scale.
1This table does not list all studies.